Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Adding an FAQ on recommended range of dataset size for cudf.pandas #16869

Closed
wants to merge 7 commits into from
Closed

Adding an FAQ on recommended range of dataset size for cudf.pandas #16869

Changes from all commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
5a3e609
added an faq on recommended range of dataset size
Sep 22, 2024
e8dfdf0
added an faq for recommended range of dataset size
Sep 22, 2024
188c0b9
adding placeholder for experimental details
Sep 23, 2024
a335016
Modified recommended dataset sizes in the faq
singhmanas1 Sep 23, 2024
de1ce2a
Added link to 'how-it-works' page
singhmanas1 Sep 23, 2024
900c4cd
Merge branch 'branch-24.10' into feat/faq_cudf_pandas
galipremsagar Sep 24, 2024
4122493
Merge branch 'branch-24.10' into feat/faq_cudf_pandas
galipremsagar Sep 25, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 11 additions & 0 deletions docs/cudf/source/cudf_pandas/faq.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -194,3 +194,14 @@ for testing or benchmarking purposes. To do so, set the
```bash
CUDF_PANDAS_FALLBACK_MODE=1 python -m cudf.pandas some_script.py
```

## What is the recommended range of dataset size that I can process using `cudf.pandas`?

`cudf.pandas` can process a wide range of dataset size. As a
_very rough_ rule of thumb,`cudf.pandas` shines on workflows with more than
10,000 - 100,000 rows of data, depending on the algorithms, data types, and other factors.
Below this range, workflows might execute slower on GPU than CPU because of the
cost of data transfers. With [managed memory pool and managed memory prefetching enabled in cudf
by default](how-it-works.md), you can process datasets larger than GPU memory and up to a theoretical
limit of the combined CPU and GPU memory size. However, note that the
best performance with large data sizes can be data and workflow dependent.
Loading